Machine Learning Feature Engineering Pipeline

Develop a MySQL-based feature engineering pipeline that can automatically generate, transform, and select machine learning features directly within the database. The solution must support multiple feature generation techniques, handle feature scaling, detect multicollinearity, and provide statistical metadata about generated features. Include performance optimization techniques to ensure the pipeline can process large datasets efficiently.